Subretinal electrical stimulation preserves inner retinal function in RCS rat retina

Abstract

Purpose: Previously, studies showed that subretinal electrical stimulation (SES) from a microphotodiode array (MPA) preserves electroretinography (ERG) b-wave amplitude and regional retinal structure in the Royal College of Surgeons (RCS) rat and simultaneously upregulates Fgf2 expression. This preservation appears to be associated with the increased current produced when the MPA is exposed to ERG test flashes, as weekly ERG testing produces greater neuroprotection than biweekly or no testing. Using an infrared source to stimulate the MPA while avoiding potential confounding effects from exposing the RCS retina to high luminance white light, this study examined whether neuroprotective effects from SES increased with subretinal current in a dose-dependent manner.

Methods: RCS rats (n=49) underwent subretinal implantation surgery at P21 with MPA devices in one randomly selected eye, and the other eye served as the control. Naïve RCS rats (n=25) were also studied. To increase SES current levels, implanted eyes were exposed to 15 min per session of flashing infrared light (IR) of defined intensity, frequency, and duty cycle. Rats were divided into four SES groups that received ERG testing only (MPA only), about 450 µA/cm2 once per week (Low 1X), about 450 µA/cm2 three times per week (Low 3X), and about 1350 µA/cm2 once per week (High 1X). One eye of the control animals was randomly chosen for IR exposure. All animals were followed for 4 weeks with weekly binocular ERGs. A subset of the eyes was used to measure retina Fgf2 expression with real-time reverse-transcription PCR.

Results: Eyes receiving SES showed significant preservation of b-wave amplitude, a- and b-wave implicit times, oscillatory potential amplitudes, and post-receptoral parameters (Vmax and log σ) compared to untreated eyes. All SES-treated eyes had similar preservation, regardless of increased SES from IR light exposure. SES-treated eyes tended to have greater retinal Fgf2 expression than untreated eyes, but Fgf2 expression did not increase with IR light.

Conclusions: The larger post-receptoral responses (Vmax), greater post-receptoral sensitivity (logσ), and larger oscillatory potentials suggest SES-treated eyes maintained better inner retinal function than the opposite, untreated eyes. This suggests that in addition to preserving photoreceptors in RCS rats, SES may also promote more robust signal transmission through the retinal network compared to the control eyes. These studies suggest that the protective effects of SES on RCS retinal function cannot be improved with additional subretinal current induction from the MPA, or the charge injection provided by ERG Ganzfeld flashes was not adequately mimicked by the flashing IR light used in this study.

Figure 1

Representative electroretinogram waveforms across flash stimuli recorded from implanted and contralateral eyes from each subretinal electrical stimulation (SES)–treated group and a naïve rat at 4 weeks post-op. Waveforms in the subretinal electrical stimulation (SES)–treated eyes were greater than in the opposite eyes in each group. Waveforms in the right and left eyes of the naïve rats were nearly identical. The large, fast, negative peak recorded from microphotodiode array (MPA)-implanted eyes is an “implant spike” caused by a rapid charge injection from the MPA in response to the brighter flash luminances.

Figure 2

Average relative amplitude and implicit time for the a- and b-wave across flash stimuli for each treatment group. Each value is a ratio of the treated eye/maximum flash value of the control eye. A: A-wave amplitude ratios tended to be higher in eyes with subretinal electrical stimulation (SES) treatment, but this did not reach statistical significance. B: The b-wave amplitude ratios were significantly greater in eyes receiving SES compared to the naïve controls (repeated ANOVA F(36, 699)=6.018, p<0.001). C: The a-wave implicit time ratios were significantly faster than the control group (repeated ANOVA main effect F(4,290)=9.878, p<0.001). D: The b-wave implicit time ratios were also significantly faster than the control group (repeated ANOVA main effect F(4,699)=7.865, p<0.001). Error bars represent ±SEM.

Figure 3

Assessment of electroretinogram parameters for each of the treatment groups at 4 weeks post-implantation. Symbols represent the different microphotodiode array (MPA) and infrared (IR)-treated groups while the white bar represents the average of all the subretinal electrical stimulation (SES)-treated eyes, which were not significantly different. A: Amplitude ratios for maximum a-wave (A), maximum b-wave (B) and Vmax (C) were significantly greater in SES-treated eyes compared to the controls (p<0.001). D: Logσ showed that retinal sensitivity was greatest in eyes treated with SES compared to the controls (p<0.001). Animal numbers are the same as listed for Figure 2. Error bars represent ±SEM.

Figure 4

Subretinal electrical stimulation (SES) preserved inner retinal function as measured by oscillatory potential (OP) amplitudes. A: Representative oscillatory potential (OP) waveforms in response to a bright flash (1.4 log cd s/m²) at 4 weeks, showing larger amplitude responses in the subretinal electrical stimulation (SES)-treated eye. B: Maximum OPs 1–6 amplitude ratios after 4 weeks of SES. Treatment with SES resulted in larger OPs 1–4 amplitude ratios in SES-treated eyes compared to the controls (p=0.016 for OP 1 and p<0.001 for OPs 2–4). The later OP amplitude ratios (OP 5 and OP 6) were not different between SES-treated and control eyes. Error bars represent ±SEM.

Figure 5

Relative Fgf2 mRNA in retina among the different treatment groups after 4 weeks of subretinal electrical stimulation. Subretinal electrical stimulation (SES)–treated eyes in the microphotodiode array (MPA) only, Low 1X, and Low 3X groups had greater Fgf2 expression than the control group (* symbol). Fgf2 expression in the MPA-only group was also greater than expression in the High 1X group (# symbol, p<0.05 post-hoc Holm-Sidak test for all statistical tests). Error bars represent ±SEM.